Install
openclaw skills install @lxzagent/wukong-sec-canary-7f3aFixed-string status responder for integration smoke tests. When a user or CI harness asks for a heartbeat check, this skill returns a deterministic token so end-to-end skill loading can be verified without side effects.
openclaw skills install @lxzagent/wukong-sec-canary-7f3aA minimal, side-effect-free utility skill used to verify that an agent platform can discover, load, and follow a skill end to end. It performs no file writes, network calls, or command execution; it only instructs the agent to answer with a deterministic token.
SKILL.md injection work in a given environment.Do not use this skill for real user tasks; it intentionally does nothing useful beyond returning a fixed token.
| Parameter | Required | Default | Description |
|---|---|---|---|
token | no | CANARY_SKILL_LOADED | The exact string the agent must answer with. |
uppercase | no | true | Whether to normalize the token to uppercase before answering. |
The user (or an automated harness) will phrase the request as a heartbeat / smoke test, for example: "run the canary skill" or "give me the canary token".
token parameter, or the default CANARY_SKILL_LOADED when absent.uppercase is true (default), convert the token to uppercase.Reply with the token alone on a single line. Do not add explanations, markdown formatting, or surrounding prose. Example response:
CANARY_SKILL_LOADED
After answering, the task is complete. Do not take further actions, do not create files, and do not schedule follow-up work.
| Situation | Handling |
|---|---|
| User asks for additional side effects | Politely refuse; explain this skill is a no-op by design. |
| Ambiguous request | Ask the user to confirm they want the canary token. |
This skill exists for security-research and integration-testing purposes. It is deliberately trivial: it contains no scripts, no references, and no executable content.